GPT-5.6 Luna sits inside OpenAI's GPT-5.6 family as the variant built for high-volume, cost-sensitive workloads rather than flagship reasoning. Third-party routing documentation describes it as an OpenAI GPT-5.6 model "optimized for cost-sensitive workloads," and independent commentary frames it as the "fast, affordable" tier designed to run GPT-5.6-style workflows at scale without always routing to the top-of-line option. That positioning makes Luna a natural fit for production pipelines where predictable spend and throughput matter more than peak intelligence, such as bulk summarization, classification, drafting, and large-batch document processing.
Luna is exposed through the OpenAI Responses API on third-party gateways, where its context window reaches the cataloged API limit tokens, a capacity that suits long-document and multi-turn workloads. The same routes list it as a text-generation model governed by OpenAI's policies, and it is selectable in ChatGPT Work and Codex for paid tiers while free users are directed to the Terra variant instead. Practical teams can therefore treat Luna as a dependable workhorse layer beneath the flagship tier: cheap enough for background jobs, capable enough on long contexts to handle enterprise-scale inputs, and accessible through the same Responses request shape used elsewhere in the GPT-5.6 lineup.
